Nanoscale Filament Evolution and Local Resistive Switching in Substoichiometric Yttrium Oxide Thin Films
Abstract Scaling resistive random-access memory (RRAM) toward ultradense crossbar arrays for high-density storage and neuromorphic computing requires precise nanoscale control of conductive filaments for reliable multilevel operation. Here we directly visualize and modulate nanoscale filament evolution in substoichiometric Y2O3–x thin films using high-resolution in situ conductive atomic force microscopy (c-AFM) with controlled current compliance. The switching threshold systematically increases with increasing oxygen stoichiometry, while filament evolution is continuously modulated by varying the applied bias. The initially formed larger filaments persist and dominate subsequent switching events during multicycle measurements. Using the c-AFM tip as a nanoscale mobile top electrode, single-point switching reveals localized switching with a lateral influence of approximately 20 nm, defining a crosstalk length scale. These results directly link oxygen stoichiometry, microstructure, filament evolution, and switching length scales, providing new insight into nanoscale filamentary switching and physical guidelines for scaling polycrystalline memristors toward ultradense memory and neuromorphic applications.
Authors
- Eszter Piros (ORCID: https://orcid.org/0000-0001-8714-9059)
- Taewook Kim (ORCID: https://orcid.org/0000-0002-3095-0053)
- Yen‐Po Liu (ORCID: https://orcid.org/0000-0003-0144-0991)
- Philipp Schreyer (ORCID: https://orcid.org/0000-0002-7200-3855)
- Lambert Alff (ORCID: https://orcid.org/0000-0001-8185-4275)
- Regina Dittmann (ORCID: https://orcid.org/0000-0003-1886-1864)
- Stefan Wiefels (ORCID: https://orcid.org/0000-0003-2820-9677)
- Yu Duan (ORCID: https://orcid.org/0009-0006-3133-403X)
- Yingxin Li (ORCID: https://orcid.org/0009-0002-1134-1794)
- Erkai Wang
- Alexey Arzumanov
- Luisa Bayer
Institutions
- Forschungszentrum Jülich (DE)
- Technische Universität Darmstadt (DE)
Publication Details
- Journal
- Nano Letters
- Published
- 2026-09-28
- DOI
- https://doi.org/10.1021/acs.nanolett.6c04262
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00